Love it! My Mom is German. I grew up around it. I just have not used it for 37 years. And technical German was not really used around the house. I included a couple more articles on calibration.
Yes nothing is idiot proof, they always invent better idiots.
Same for AI human nature is not exactly predictable is it
Mark
I see a couple of options to control the 3 steppers:
Does anyone have any good advise? A list of pros and cons? A well contemplated solution?
- tic, easy but gets expensive for 3 axis
- arduino + cnc shield
- 3d printer board + marlin
- 3d pritner board + raspberry pi + klipper
- ?
Does anyone have any good advise? A list of pros and cons? A well contemplated solution?
depending on how much interest there is, I am more than happy to make a proper PCB.
All that being said, for first prototype, just get a ready board first.
If memory serves well, those Arduino CNC boards are easy to use/hack/abuse.
Actually the “shield “ portion has other options as well.
Why: Octave has Arduino support; it also doesn’t hurt that Arduino is cheap.
Microstepping and some other tweaks have limited noise considerably. However, there is a position dependent high pitched noise when they are stationary. For now I also disabled the motors during the measurements.
They are quite a bit more expensive, but they solve a problem...
https://www.aliexpress.com/item/1005002107565356.html?spm=a2g0o.detail.pcDetailBottomMoreOtherSeller.13.6a45iNW6iNW6hv&gps-id=pcDetailBottomMoreOtherSeller&scm=1007.40000.326746.0&scm_id=1007.40000.326746.0&scm-url=1007.40000.326746.0&pvid=588df23a-cc36-4e54-9e94-96e92562d274&_t=gps-idcDetailBottomMoreOtherSeller,scm-url:1007.40000.326746.0,pvid:588df23a-cc36-4e54-9e94-96e92562d274,tpp_buckets:668#2846#8107#66&pdp_npi=4@dis!CAD!42.88!42.88!!!30.88!30.88!@2103205217109651831928111ecc7e!12000018763258558!rec!CA!122373966!&utparam-url=scenecDetailBottomMoreOtherSeller|query_from:
A method to be able to know our angular position. If we use a motor with an encoder we are very far from the actual mic position. If we want a few mm positional accuracy it might be smart to have some reasonably priced rotary encoder for rotational position.
Asking for comments. Shoot it down? I'm thinking out loud here having worked on many different types of machines. There are losses due to flexture in the solid moving arm, loose fittings, gear meshing clearances that are required. They all add up quickly to a system where you really do not have the real world resolution that you think you have.
Mark
cDetailBottomMoreOtherSeller,scm-url:1007.40000.326746.0,pvid:588df23a-cc36-4e54-9e94-96e92562d274,tpp_buckets:668#2846#8107#66&pdp_npi=4@dis!CAD!42.88!42.88!!!30.88!30.88!@2103205217109651831928111ecc7e!12000018763258558!rec!CA!122373966!&utparam-url=scenecDetailBottomMoreOtherSeller|query_from:A method to be able to know our angular position. If we use a motor with an encoder we are very far from the actual mic position. If we want a few mm positional accuracy it might be smart to have some reasonably priced rotary encoder for rotational position.
Asking for comments. Shoot it down? I'm thinking out loud here having worked on many different types of machines. There are losses due to flexture in the solid moving arm, loose fittings, gear meshing clearances that are required. They all add up quickly to a system where you really do not have the real world resolution that you think you have.
And they are quiet. Steppers holding position are not quiet. They whine. A brake will buzz a bit to.
Mark
Would solve some issues with people on different platforms that have to install and setup stuff manually. We could just provide an image for the raspberry.
But on the other hand, sound a bit like gold plating while we're still wrestling in the mud.
The mechanical side is doing nothing if we don't have something worked out to move it.
https://www.aliexpress.com/item/1005002107565356.html?spm=a2g0o.detail.pcDetailBottomMoreOtherSeller.13.6a45iNW6iNW6hv&gps-id=pcDetailBottomMoreOtherSeller&scm=1007.40000.326746.0&scm_id=1007.40000.326746.0&scm-url=1007.40000.326746.0&pvid=588df23a-cc36-4e54-9e94-96e92562d274&_t=gps-id
A method to be able to know our angular position. If we use a motor with an encoder we are very far from the actual mic position. If we want a few mm positional accuracy it might be smart to have some reasonably priced rotary encoder for rotational position.
Asking for comments. Shoot it down? I'm thinking out loud here having worked on many different types of machines. There are losses due to flexture in the solid moving arm, loose fittings, gear meshing clearances that are required. They all add up quickly to a system where you really do not have the real world resolution that you think you have.
And they are quiet. Steppers holding position are not quiet. They whine. A brake will buzz a bit to.
Mark
Actual position is not that important. In the Klippel document attached a few posts back, they optimize the source location and orientation for each frequency.
I would say we don't need an encoder as long as the stepper is not skipping steps.
Then a hybrid stepper would cover that problem nicely.
Mark
As an aside, there is an enormous amount of making up for imprecise positioning in this "super precise" system. I am learning so much here. It is wonderful. I really appreciate everyone's comments and ideas.
The klippel rotates the whole measuring arm. Then is radially moves and finally is moves the microphone up and down.
In ordinary speaker impulse measurements I think I see the influence of the mic stand in the measurements. @fluid pointed me to this:
For SFS we need to measure at two distances from the speaker. If we change the 'order of motion' to rotation, up/down and finally radially, most of the reflections from measuring device will be the same for the 2 measurements at two (radial) distances. It feels like that would give SFS the best opportunity to even calculate out the reflections from the measuring device, whereas with the Klippel NFS it can't do that as the reflections of the vertical boom change. Makes sense?
For rigidity is has both pros and cons I think.
In ordinary speaker impulse measurements I think I see the influence of the mic stand in the measurements. @fluid pointed me to this:
For SFS we need to measure at two distances from the speaker. If we change the 'order of motion' to rotation, up/down and finally radially, most of the reflections from measuring device will be the same for the 2 measurements at two (radial) distances. It feels like that would give SFS the best opportunity to even calculate out the reflections from the measuring device, whereas with the Klippel NFS it can't do that as the reflections of the vertical boom change. Makes sense?
For rigidity is has both pros and cons I think.
Yes exactly. The mic and the mic stand reflections are always present in a normal measurement.
What are we rotating? Mic Boom like Klippel or Loudspeaker?
Mark
And if they are not the same in both measurements, we can't calculate out the reflections.
Mic boom. Otherwise we can't calculate out the room reflections.
That makes sense. The two stationary things are the loudspeaker and the room.
Mark
I find that picture rather strange.
The mechanical momentum is the biggest by (a), followed by (b) and very little by (c), since the mass is basically at the mechanical reference point.
So unless this picture and experiment was taken on a different planet, but I find the results very strange to say the least?
Unless we are looking at reflections from the beam instead of vibrations, but it's a bit unclear at this point.